Motor stall damage

I’ve been wondering recently about the physics of motors burning out/getting damaged by current draw. Particularly, I’ve been wondering what physically happens to a motor as it burns out, when motors will start to get damaged from stalling, and how to check if a motor has been damaged by stalling.

I’ve done some research about this, and I have some guesses as to what the answers are, but I’m not sure if I’m right or if there’s anything that I’m missing. Here’s what I know:

What physically happens to a motor as it burns out?

I couldn’t find a lot of information on this. Basically, my impression is that it’s drawing the maximum amount of current, which generates heat that damages the motor. But what part of the motor is damaged by the heat? The coils? The brushes?

The reason that it draws the most current when stalling is, as far as I can figure out, related to the back EMF - when the motor is not moving, it’s not generating any back EMF. When this happens, the force applied from the current going into the motor, and thus the motor draws the full current, since it’s not being counter-acted by the back EMF. Is this correct? Also, when that’s the case, what’s determining what the stall current is? (I know that I can find this on the datasheet, but I’m wondering what the actual property of the motor is that determines this is).

The last reason that I can find about why it gets damaged is because, in motors that have fans in them (such as the 775pro), the fan isn’t spinning during stall, thus there is only passive heat dissipation.

When will motors start to get damaged from stalling?

I’m really not sure about this - this question is the main reason that I made this post. I’ve heard that motors will start to have their efficiency permanently reduced. But when does that start to happen? I’m assuming that the answer to this is related to the answer to the above question.

How to check if a motor has been damaged by stalling?

The only way that I know to do this is to measure the resistance of the motor, but I’m not really sure how to interpret this data. For example, on a 775pro, I’d expect 0.4Ω, but what if it isn’t? For example, I’ve seen 775pros with 1Ω of internal resistance, which still appeared to be working. Does that mean that it’s damaged? If so, how much will that effect the output of the motor? Is there any way to find what the expected internal resistance of a motor is, aside from measuring a known good motor?

Are there any other good debugging steps for potentially damaged motors?

Several things can happen. The basics are this. Some motors have windings made from thin CU wires. These wires have a thin varnish on them that acts like an insulator so the wires function as a big coil and not a big blob of CU. If you over heat a motor the varnish can melt or fail. Then the coils short against each other effectively making the coil shorter or not a coil at all.

The other issue is where the stator and armature are connected. The brushes or slip rings allow current flow the armature while is spinning. The contact points are relatively small yet carry the full motor current. When the motor stalls the brushes and contacts get pitted, overheated, and over time destroyed.

It all depends on the motor type but those are the basic concepts.

1 Like

For starters, you’re going to have difficulty measuring resistance in that range accurately with a regular multimeter.

The two primary failures you’re going to see (before full failures) is partial melting of the insulation on the copper, shorting out some (but not all) of the coils which will change the torque and voltage constants (not for the better!).

You’ll also see the brush assemblies gets damaged by high temps, and the magnitude of this depends on their type and quality.

One of the two affects might happen fully before the other.

Vexpro has good data on their site about how long the various rotors last at locked rotor stall in 2V increments (and you can see resultant current in their graphs). This neglects the fan helping as it’s not moving, but it provides a nice worst case baseline.

Unless a motor is starting to be appreciably damaged and/or you don’t have good measuring equipment a dyno test will better identify a failing motor because it’s easy to do while installed in the robot using existing functionality.

Writing some test code to take advantage of the current measurement in PDP/talon and the velocity feedback from any sensored system should be able to easily characterize a systems open loop response to a given voltage (in steady state RPM and current). Be aware that you also might pick up mechanical change/failure in this test.

1 Like

I’m very interested in this as well. We currently believe that an issue we saw in a particular match was caused by this. Gradual build up of damage from minuscule stalling over multiple days. If this was the cause of the problem, it should be very easy to monitor and fix assuming there is measurable change in resistance or performance as the motor degrades.

My above post was purely related to thermal failure, but you’ll also see a slow decline in motor performance in time as the brushes wear out and the motor becomes more filled with debris. You’ll also potentially see an initial increase in performance as brushes wear in.

Take a look at this page which describes brushed DC motor equations.

Also an old CD thread with info:

When a motor is at stall, Power into the motor = V*I. Power out of the motor = 0. All of that power is converted to heat, and there is lots of it. When there is lots of heat with nowhere to go, things start to melt.

Hmm. How accurate do you need to be with this? The fluke multimeter that I usually use has a precision of 0.1Ω, and I regularly see 0.4Ω when I test 775s. Is the resistance change when the motor is damaged less than 0.1Ω? I recently saw a motor that was reading 1Ω, but seemed to be running fine, which is why I assumed that you’d see a larger change in resistance (although I was measuring from the anderson, so there may have been a bad solder joint).

Ok, that’s pretty much what I assumed. Do you know if there’s data on when these failures will start happening? I’ve seen the VEX locked rotor stall torque vs time charts, but I don’t know when actual damage starts occurring.

Isn’t some of the power being used as torque being applied to the mechanism? (Until, of course, the motor completely dies).

Isn’t some of the power being used as torque being applied to the mechanism? (Until, of course, the motor completely dies).

No, Power is work over time. The type of energy output by a motor is mechanical. If there is no movement, then there is no work being done on the system attached. All the work is being done on the motor in the form of heat.

In electricity you can have a voltage with no current flow, thus no workign being done. In mechanics you can have a force but no movement, so no work.

Smell it! The more it smells like smoke the more likely it needs replacement…